dissenting:
It is the law of our circuit that when a district court denies discovery, the party complaining of error must show actual and substantial prejudice to obtain a reversal on appeal. See Sablan v. Dep’t of Finance of N. Mariana Islands, 856 F.2d 1317, 1321 (9th Cir.1988) (a court’s “decision to deny discovery will not be disturbed except upon the clearest showing that denial of discovery results in actual and substantial prejudice to the complaining litigant.”) (inner quotation marks omitted); Butcher’s Union Local No. 498 v. SDC Investment, Inc., 788 F.2d 535, 540 (9th Cir.1986); Data Disc, Inc. v. Systems Technology Associates, Inc., 557 F.2d 1280, 1285 n. 1 (9th Cir.1977).1 The majority disregards circuit law and holds that a .denial of “meaningful discovery” is a structural error that is per se harmful.2
*949The majority goes astray in its analysis by focussing on the impact of a denial of-discovery on the discovery process itself rather than its impact on the outcome of trial. The following statement makes this clear: '“Although we do not know what kind of case Martel actually put on at trial, we do know that his preparation for that trial was manifestly prejudiced.” Majority Opinion at 945-46. It is also settled law, however, that prejudice must' be measured in terms of the likelihood that the alleged error changed the result of the trial. See Chapman v. California, 386 U.S. 18, 22, 87 S.Ct. 824, 827, 17 L.Ed.2d 705 (1966) (focussing on the “likelihood of [the error] having changed the result of the trial”). Unless we know “what kind of case Martel actually put on at trial”, we cannot judge the likelihood that further discovery by Martel would have affected the outcome of that trial.
The reason we do not know what kind of case Martel put on at trial is that he has made no effort to tell us. For all .we know, Martel’s case was so weak that it would not have survived a summary judgment motion had there been one. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 249-50, 106 S.Ct. 2505, 2510-11, 91 L.Ed.2d 202 (1986) (“[T]here is no issue for trial unless there is sufficient evidence favoring the nonmoving party for a jury to return a verdict for that party_ If the evidence is merely color-able ... or is not significantly probative ... summary judgment may be granted.”) (inner quotations omitted). For all we know, Martel’s theory of the case is facially implausible and his evidence completely lacking in probative value. And for all we know, the testimony of the officers involved may have been completely plausible and fully supported by the physical and documentary evidence.
On appeal, the burden is on Martel to make a showing of a reasonable likelihood that the denial of further discovery affected the outcome of the trial. He cannot carry that burden without telling us at least something about what happened at trial. But he tells us nothing, not what the testimony was, on what points it was in conflict, or how additional discovery might have been helpful in presenting the case to the jury in a light more favorable to Martel on the disputed issues. By failing to tell us anything about the trial, or to provide us with a trial transcript or even portions thereof, Martel fails to carry his burden of showing a likelihood that the denial of further discovery would have changed the result of the trial; The majority is dead right that “we do not know what kind of case Martel actually put on at trial.” Martel has not told us.
Nonetheless, the majority excuses Martel from his burden of showing that the outcome of trial was likely affected by the denial of discovery because the majority believes that “it is both unfair and pointless to require a specific showing of the impact the district court’s error had on the outcome of the proceeding.” Majority Opinion at 944-45. Even if the prejudice requirement were not the law of the circuit, I would disagree with the majority that we should adopt a structural rule dispensing with the requirement of showing trial prejudice in cases of denial of discovery.
“Structural” rules that presume prejudice from trial court error are the exception, not the rule. Generally, parties are required to show that trial court error was in fact prejudicial.3 I see no justification for creating an exception to the general rule for what the majority calls a denial of “meaningful discovery.”4 It is not unduly burdensome to require a party seeking reversal for a denial of discovery to tell us at least something about *950the trial issues and evidence that would help us to make a judgment on the likelihood that additional discovery would have affected the outcome of the trial. Litigators routinely dissect a trial to find a basis for arguing that trial court error was prejudicial. In this case, for example, Martel may have been able to carry his burden merely by showing that the case turned on a battle of credibility between his witnesses and the officers, that the testimony on both sides was reasonably credible on its face, and that the outcome could easily have been different if he had been given the opportunity to probe for possible inconsistencies in the officers’ testimony in the informal, relatively risk-free setting of a deposition outside the presence of the jury. For our part, we routinely review such claims of trial prejudice despite their inherently speculative nature. But we are not expected to decide such issues without help from the party asserting the claims. In this case, Martel has provided us with no help whatsoever. In sum, I see no basis in reason, experience, precedent, or the circumstances of this particular case for creating a discovery exception to the general rule requiring a showing of a likelihood that trial court error affected the outcome of the trial.
Finally, the majority’s reliance on United States v. Lane, 765 F.2d 1376 (9th Cir.1985), as authority for relieving Martel of his burden of showing prejudice in terms of the outcome of the trial is misplaced. See Majority Opinion at 945. In Lane, we held that Lane failed to show actual prejudice because he “ha[d] not specified which witnesses his defense attorney would have interviewed, which documents and exhibits he was unable to examine, or what defenses he might have explored.” Id. The majority cites Lane as authority that Martel has carried his burden of showing prejudice simply by identifying witnesses he wanted to depose and documents he wanted to examine. But this reading turns Lane on its head. Lane merely holds that in failing to specify any witnesses, documents or theories he would have pursued on discovery, he failed to make the required showing of prejudice. It cannot be read as authority for the logically different proposition that Martel carried his burden of showing prejudice merely by identifying some witnesses he wanted to depose and some documents he wanted to examine. In other words, Lane does not inform the question whether Martel made a sufficient showing of prejudice to get a new trial; it tells us only that Lane fell short in identifying no witnesses or documents whatsoever.
In sum, Martel has failed to carry his burden of showing trial prejudice because he has faded to make any showing that the outcome might have been different had he been given a further opportunity for discovery. He tells us nothing about the trial except that the jury returned a defense verdict. This is not enough. Nonetheless, the majority bails him out by adopting a new structural rule that irrebuttably presumes prejudice when there is a denial of “meaningful discovery” — whatever that means — a rule that not only conflicts with circuit precedent, but is unsound as a basis for reversing judgments entered following trial.5 I dissent.
. In this case, as in Sabían, “a continuance was sought for the purposes of conducting discovery, and hence its denial was effectively a denial of discovery.” Id. at 1321. See also United States v. Lane, 765 F.2d 1376, 1379 (9th Cir.1985) (movant must show actual prejudice before a denial of a motion for continuance will be reversed); United States v. Mitchell, 744 F.2d 701, 705 (9th Cir.1984) (same).
. Although the majority does not expressly use the term structural error, it is clear that the majority adopts precisely this kind of rule. The majority writes: "it is both unfair and pointless to require a specific showing of the impact the district court's error had on the outcome of the proceeding.” Majority Opinion at 944-45. As authority for this proposition, it cites Hays v. *949Arave, 977 F.2d 475 (9th Cir.1992), which is a structural error case. See id. at 482 (unconstitutional in absentia sentencing of a criminal defendant is structural error).
. See Fed.R.Civ.P. 61 ("[N]o error or defect in any ruling ... is ground for ... setting aside a verdict ... unless refusal to take such action appears to the court inconsistent with substantial justice. The court at every stage of the proceeding must disregard any error or defect in the proceeding which does not affect the substantial rights of the parties.”).
. The majority makes no attempt to give content to its "meaningful discovery” standard. Instead, it muddies up the legal waters by employing an elastic, vague, and ultimately meaningless phrase that will be the source of vexatious litigation in the future.
. Because I believe Martel has failed to make the required showing of actual prejudice, I do not address whether the district court abused its discretion in denying Martel's motion for a continuance.